After years of studying baccarat I've come to the conclusion that the only way to win at this game is by properly assessing the number and distribution of BP shifts per each shoe.

Each shoe features an average probability to get this or that, first we should restrict the number of such this and that.

We cannot care less about long streaks, they are the best way amateurs try to get a profit by.
Long streaks or homogeneous patterns or predominance factors are just post hoc findings.
They will come at the right time or not.
As players we are compelled to restrict the "right time" within "now" or "very shortly" terms. A thing that the random world is laughing at.

Baccarat shifts move more from P to B than from B to P but itlr (and intermediate terms too) the number of BP shifts is equal, actually is slightly oriented to get more shifts than a perfect 50/50 proposition dictates.
Such conclusion came from testing millions of shoes.

We know that in a 50/50 perfect proposition itlr the number of singles and doubles will be very close to 75%.
Of course B singles and B doubles itlr will get a lesser amount than 75% compensated by the P outcomes.

Really?

No way.

A large sample study about sd values calculated on 3+ or 4+ streaks formed on both sides tell us that the number of shoes featuring a low than average number of such streaks will outbalance the number of shoes featuring a higher than average number of those streaks.

Of course along any shoe the probability to get an higher than average number of such streaks will come out more often than not whether those streaks had come out by a higher pace than expected.

Why?
Because actual card distribution which produced such "unexpected" long streaks has consumed space to get the more likely shifthing mood.
And the same is true regarding more likely situations.

The effect is working on every shoe dealt in the universe but it's more likely to happen when we have reasons to think that the distribution won't be perfectly random.

And in the actual bac world we're not betting against pc distributions.

as.

That's nice Al! :-)

Every bac player should think about this.

An almost 50/50 slight dependent distribution will form certain unidirectional results for some time.
Not every shoe will be so polarized but most part of shoes will present such feature.

Of course B/P gaps or other too variance affected outcomes won't give us many of hint.

Say we have won (or lost) 5-6 flat betting units so far (half or more of the actual shoe).

Now, are we going to bet toward the deviations happened so far or to get a kind of balancement situations?

Mathematicians will say it doesn't matter which kind of direction we'll take.
But they are wrong.

Since we cannot guess hands, the more we'll play the higher will be the probability to get opposite results already gotten, thus taking an unidrectional strategy sooner or later (I'm speaking about 4-5 shoes not more) will surely fail.

After 5 shoes played, the probability to be ahead is just 20% or so.
Thus, after 5 shoes, the probability to get some opposite results is very high at some point.

Guess what?

Balancements are going to appear when deviation situations will cross a 2 or 3 cutoff point.
Nobody wants to play a 2 or 3 or higher negative proposition hoping to get a single positive outcome, therefore a smart player would know when things are going to change.

I mean that it's not wrong to follow a positive pattern providing to have secured a profit, but it's highely recommended to stop the betting toward positive deviations whenever a 2 or 3 negative step is going to show up.

More on that tomorrow

as.
 



   









 

There's an interesting line of thought suggesting that the first half or 2/3 results of a shoe will slightly affect the probability of the remaining outcomes.

More later.

as.








   

Quote from: Jimske on November 28, 2018, 09:34:34 PM
Right, very rare.  But lesser productions of P single not so rare and is why I use the 1's as a key to help determine the bias.  We can use a bet placement that exploits this.  Any positive expectation (weighted count, regression to mean, etc.) has escaped me.  Enter the "educated" guess.
Assuming you or anyone has identified such it no doubt requires a potential long wait.  Waiting even for Sputnik's Ecart still produces the problem with variance due to the low odds.  We'd have to be in the neighborhood of 75%+ prediction to make it practical.

Do you disagree?

Nope, you are completely right.

We need some time to consistently win in the same way casinos need time to let the math edge or variance destroying players' bankrolls.
Only recreational players try to win every or almost every single situation dealt and that's not possible.

Imo, time and "space" are the two most important factors when playing baccarat.
Time allows math and variance to show up and space is the physical factor that works by time frames.

Spaces can be reduced by bankroll managements, nothing we can do about time but waiting.

as. 

Although your "sections and turning points" post is very valuable, we can't forget that power of math and statistics.

For example, we know for sure that somewhere banker will produce a streak or a given amount of streaks. There's no way that a given shoe will produce just one B streak, but we surely know that very rare given shoes won't produce a P single.
The answer is easy: streaks consume a well more room than singles but B side is more prone to produce streaks as it's more likely to happen.

Say we want to bet a very complicated and hyper selected strategic plan oriented to lose only when a shoe won't present a P single AND at least two B streaks.
We won't lose by 1 billion accuracy.

as.   

   

Thanks Jimskie!!!

After all singles provide less issues than streaks. A single is a single, streaks are of many kinds.
I mean that if we hope to get an initial streak, we should be happy to win the first hand. If we keep flat betting the streak, we need at least a streak of three to be ahead providing we continue to wager the streak until it stops.

Third card flow.

Everybody has noticed that more often than not  third cards greatly helps or totally not the player side. Sometimes miracles happen but they do not alter too much the shoe texture.
The same, this time more easy to understand, when player is standing and banker is drawing a third card. Obviously we do not want to find ourselves in such situation.

Remember that from a general point of view, a third card is a helpful card more often than not.
Asians like to shout "monkey" when they are betting a possible not drawing banker and waiting for the third card to show up, but they forget that the probability to get a monkey is "just" 30.76%.
Thus nearly 70% of the times they are dog to see what they wish to see.

Of course a third card could be an 8 or 9 or an high card damaging the Player or Banker hand. Or this card is going to make an unbeatable or hugely favorite P hand. Banker has the advantage to act after Player but it has no gain to draw if P is standing.

So how to consider a third card as good or bad from a player or banker perspective?

First, when playing baccarat we shouldn't want to see a third card falling on the side we are betting into.

If it happens we are losing money, except when we catch the quite rare pure drawing-drawing situations and we're betting P side.

When betting P side we want a standing point, when we are betting B side we want P drawing and B having any point different from 0, 1 or 2.

Nonetheless, third card has a sort of natural flow in direction to either help or not the side we are betting into.

Most people don't focus enough on third card nature as a large part of outcomes is decided by first four initial cards.

More later.

as.












 




 







     

Al, of course your post provides important informations

Almost all the time, people cannot cool off or stop playing and thus the disaster will strike.  Probably 99.98 to 99.99.9% of the times players win.  IMO and Experience.


This part is of extraordinary importance.

Bac players lose a lot more than expected by math as they want to recover too rapidly.
Or simply they want to recover trying to guess the unguessable.
They start betting side bets when losing and actually they should do right the opposite.

Simply put, people win less on positive patterns and lose more on negative ones.
In reality the sum is always zero (plus the vig burden).

Back to topics.

Every random game is formed by streaks and singles, the same about baccarat regarding the quality of the hands (not the results).
The exact points evaluation don't add anything worth of it; it's too complex and quite unreliable for the volatlity impact.

People emphasizes positively or negatively when "miracles" decide the hand outcome.
They shout with joy when their bet on player 4 catches a third card 7 and banker with 5 catches another 5. They win with 1 vs zero but they should be unhappy as they have totally mistaken which side to bet on. In fact P 4 - B 5 is one of the best opportunities to get an advantage (on B, of course).
Miracles (there are tons of them) are just a very small part of the total happenings at baccarat. And they tend to alter the general flow of probabilities very rarely.
They seem to do but they do not. Forget the hands won or lost by miracles. They tend to confuse.

Regarding the quality of initial 4-card hands we have four main fields to register singles or streaks at either side:

natural/standing points  -one side
drawing points  -one side
P drawing/ B standing 3-4-5-6 (asym hand) -both sides
P drawing/ B drawing

The very best player in the universe is the person capable to get the P draw/ B standing 3-4-5-6 situation at least one time over 8-9 attempts (mathematically it's about 1/11). Naturally he will bet always Banker.

The second best player in the world is the player capable to get a drawing hand on P side at least 73-75% of the times. Again he's going to wager Banker.

The third best player is the Player bettor capable to get at least 45% of the time a standing/natural hand.

Period.

You'll notice that the worst option we could have at baccarat is to get a drawing/drawing situation. It's true that if we bet P side we are playing a zero edge game but we are in huge trouble if we are wagering B (as it's payed 0.95:1 or prone to get some taxed hands as F-7, etc).

So not everytime a P drawing hand is good for Banker. Fortunately the pure B drawing hands are few (0, 1 and 2); 3,4,5,6 elicit the precious asym situation and 7,8 and 9 stand.

Fast is fine but accuracy is everything (Wyatt Earp)

At baccarat we'll win itlr only if we are able to catch the various flows (S, D, N, etc) that advantage mathematically one side.
Easy to see that it's impossible to be accurate playing every hand or most hands. It's like to admit that we are able to catch every single flow happening at the table.

Therefore now we can rely upon two different levels of thinking and classification: results and quality of the first initial points.
Sometimes those two levels collide and those are the best opportunities to put a real wager.

By now we have talked about first four initial cards. But also the third card plays a huge, albeit secondary, role.
Next time.

as.

Quote from: Jimske on November 21, 2018, 03:02:54 AM
I use the 1's only.  If they have been average by the quarter shoe I don't expect much long streaks - until the second half when card composition often changes due to certain denomination depletion.

That's interesting.
Do you want to say something more on that?

as.

Forget the real outcomes, this is just a losers topic.
Instead consider this new way of playing the game adopting those words: chopping and streaking.

No news huh?

Ok.
Go on.

Chopping= hands that form short or no streaks on either side.
Even though it could appear as a weird assumption, this is the main feature of baccarat.
Now what's short?
Surely singles, surely doubles. Maybe triples. Stop!

If you have been reading my posts, you know that the average number of 3+ streaks happening on a single shoe is around 9.
Thus everybody here knows (not there) that the average number of 4+ streaks is around 4.5.

Streaking=hands that tend to form 3+ streaks.

Every fkn single shoe in the universe will present a given amount of chopping situations and a given amount of streaking situations.
Well, a whole streaking shoe won't be produced by any means except if one or more streaks are so long that chopping won't have any room to appear.
Conversely, if you have your last bucks to spend I reckon you'll wait some favourable situations to bet on chopping patterns as they surely as hell will happen along the way.

Curiously, chopping and streaking will appear at various degrees per every shoe dealt.
The probability of average chopping/streaking occurence knowing what happened so far in the shoe was deeply studied by your "hero" :-)

Actually an english team discovered this bac flaw several years ago. They have destroyed and are continuing to destroy many casinos. What's their strenght?
They know the shoes to attack and which shoes to let it go.
Ironically they made most money at Ritz casino in London and everybody knows the Phil Ivey story. PI wasn't payed, they did.   

So rule #1.

Not every shoe is beatable by a high degree of confidence. And they need it.

Then rule #2.

A progression is worth if applied on multiple same situations coming out on different shoes and whether such betting spots are considered worth of betting. It's just an accelerator.

Rule #3

Progressions can only work if multilayered conceived and never by a linear progression.


Ranges.

Trying to guess right every hand or most hands dealt is a silly losing operation. We simply can't.
Better to play a winning range where we need just one hand to be ahead.

Say we want to get any streak in a given spot. And say we want to apply a 1-2-4 progression.
To lose we must cross a 3 singles apparition. It could happen but by a careful assessment of what happened so far the probability may be easily lowered to 11% or so. And the average apparition is 12.5%.

The same, even better, if we want to discount the spots where a 4+ streak will happen.
Again, we just need a 14% or so probability to win.

Baccarat is beatable right on the same field the house will rely upon: tiny edges infinitely working.

I'll come back soon on more topics.

as.

Thanks Al.

Actually I probably posted too many concepts. 

Anyway, it's widely known that casinos fear heavy bettors capable to win fast and quitting the table and not common mortals wishing to be right for a long time.

Imo we do not need to join a $5000 maximum table with the purpose to bet at least $2500.
We could easily set up our minimum-maximum betting standards from 1 to 2 and strictly adhering to this plan.

In a kind of taxed coin flip game we can't hope to be right after two levels of betting.

as.   
   

Quote from: Blue_Angel on November 13, 2018, 01:01:37 PM
The most probable way to lose all your money on Baccarat is by going all in on every hand, eventually there will come 1 hand to smack your cheek!

Exactly, unless we go all-in with the maximum bet right at the start. That's why in my example I put the bankroll at $1000 with a $1000 maximum limit.

Imo this is an important concept derived from mentoring people who like to bet not less than half the maximum limit.
If we know to have at our disposal just two (50% or 100%) or three (75%) betting amounts, we'll have a clearer picture of what is going on with our selection.
Only 100% bets cannot be recovered mathematically as the vig will wipe out everything itlr, even though a good selection may hopefully invert the HE.

as.

   

Nice replies on my humble board, thanks!

For one time I'd take the opposite direction, say we want to join a table with the purpose of losing the entire money we get at the table.
What's the speedest way to lose,  say a $1000 bankroll at a $10-$1000 limit table?

Is by betting the whole $1k in just on hand? No way. Actually this is the best move we can make to win, no matter the side we'll choose to wager on.
Same about splitting our one thousand in two parts, hoping to lose two more hands than we win. It's a difficult task, but we know that sooner or later will happen, especially by the HE disadvantage.
Splitting our bankroll in more portions (3, 4 or 10) won't do the job that easy, we need to lose more hands in the process.

Of course the slowest procedure we could think of about losing our $1000 is by flat betting $10, it will take a lot of time to win such "losing reward".

Idea!
We'll bet an important part of our bankroll hoping to lose, then raising the rest part of it hoping to lose again. After all, two losses in a row comes out quite frequently along the way.
Mmmhhhhh, if we bet $300 then $700 when losing the first wager we are increasing the overall probability to be temporarily ahead of $400 and the process repeats.
Mathematically and no matter how we'll split the money wagered, the more we'll proportionally bet after a loss the higher will be the probabilities to win.

Providing finite spots.

I mean that an infinite process of raising the bets from one part increases the probability to win and on the other hand raises the risk of ruin up to the point where either betting limits or our bankroll cannot sustain the action.

We may come up to the wrong conclusion that the best way to lose is the best way to win, that is to bet a lot at the start or an important part of the bankroll, then promptly hoping to get more immediate losses (or wins).

Anything different from that will just dilute the entire process and as we well know time goes in casinos' favor.

Therefore, the more we play (hands), the more we are entitled to lose and the more we raise the wagers the more are the probabilities to lose.

I mean that the more aggressive we are playing at the start and higher will be the probability to win but itlr such procedure will surely fail; on the other hand light progressions may work just on very diluted and well calibrated spots where the probability to win (or lose) is quite restricted and whether properly assessed will sum up.

Imo, we shouldn't want to guess every single hand but just aiming to get a positive outcome on a couple (or 3) of hands serially considered.

as. 
 

You are right, no human element should be applied into the game and that's why I kept stressing about asym and sym hands, average shoe composition, lenght of certain patterns and so on.

Everything we wish to set up must be more math sensible than we can, then statistics will help us a lot (imo).

Example.
For one time we do not want to win, instead we just try to find spots where symmetrical hands will come out then always betting Player.
If in our endeavour we'll find 100% of sym spots, no matter the results, we know to play an EV=0 game.
Itlr the house will get zero from our bets.
Conversely our Banker bets will suffer a lot from such play as we'll get 0.95:1 on our winning wagers.

Naturally it's impossible to find only sym spots when betting Player, thus our hope is to reduce at most the asym hands probability on such spots, transforming the game into a lesser negative EV.

On the other hand, we do not need to spot precise asym hands as the advantage on such hands is so high when betting Banker side that we could think just in terms of range of apparition. No matter the results.

Now, asym/sym average ratio is well defined as well as its volatility (sd) with the important caveat that every single shoe is finite and card dependent.

Simplyfing (in my book I've explained everything) we should choose to put in action two distinct players playing for us, one betting B on the ranges thought to be more prone to give asym hands and the other one betting P when we think that in those spots wagered a sym hand will come out.
Of course such way of thinking needs a very diluted betting strategy.

Properly wagering in this way not only will reduce the house advantage on P hands (actually some sym hands will favor the P side for card distribution issues) but will enlarge the expectation on our B hands, hopefully inverting the house edge to our favor.

What are the statistical issues favoring us ITLR?

- There's virtually no one single shoe not forming at least one asym hand;

- Actual results of sym or asym hands don't affect our overall plan, we must think in term of EV;

- Asym hands have a general probability to come out consecutively or short gapped and a specific probability to come out depending upon the cards already removed from the shoe;

- Asym and sym hands ITLR will form polarized patterns if we split the shoe into two distinct columns (B and P).

- Sym hands are coming well more likely consecutively than singled shaped, it's up to us to ascertain the portions of the shoe when such math propensity will happen most and how long. This process needs a lot of virtaul observing.

According to those premises, we see that  consistently winning is just a long term process needing a lot of patience and observation and that it can't disjointed from a strict math foundation.

as. 
 

And I found the second bet after the first loss to be a net loss.

That's correct. And that's why we should consider certain second bets as first bets, meaning that sometimes we have to wait a fictional loss without betting.
Now our second bet becomes a first bet and it won't be followed by another "second bet".

As I've sayed numerous times here, key word at baccarat is 1. We must win the very first hand of any sequence we wish to play as it's more likely to be +1 from 0 than to get 0 from -1. 
And for that matter a series of -1 situations (from 0 to -1) won't affect the next attacks operated on the same shoe, actually they just slightly endorse the probability to lose more.

The overall amount of W and L will balance itlr but almost never in the same shoe. Thus we could use the WIAR, LIAR or anything else related to that as a termomether of the shoe water: we can't get a warm or hot water from a cold or icy water and vice versa.

Imo the important thing to remember is to let it go strong cold temperatures without betting a dime and not trying to get something from adverse conditions. And of course not trying to get too much whenever tepid or warm waters show up.

But before playing we always need to set up a precise target to look for and not only trying to adhere instinctively to what the actual shoe is producing.

as. 

Glad to hear that Sputnik!

And of course there's no need to play a lot of hands or every shoe dealt.
At baccarat you can join the game whenever you want (providing some open seats are available)

If we are driving off road in the desert and some light rain is coming, we do not know if our path will be flooded by a heavier rain. So it's better to stop our ride trying to reduce at most the very unlikely case that we'll find us stuck right in the middle.

as.